Nowadays, wind energy has drawn more and more attentions because of being a clean inexhaustible energy and the wind power penetration level is constantly increasing all over the world. However, due to the fact that wind power has characteristics such as random fluctuations, intermittence, etc., integration of a large-scale wind power might worsen the grid stability and power quality. To improve controllability of power energy output, energy storage technology is usually used in combination with the wind power generation. For example, in many applications, an energy storage system (ESS) is used to facilitate wind power to meet grid code requirements.
At the same, many control strategies for the ESS have been developed. One known strategy is based on Fuzzy-Logic based adaptive regulator (FLAR) but its membership function is hard to choose. A filter based approach can smooth the wind power output variation but it always needs a relatively high energy capacity. A method combining both the pitch angel control and the ESS can offer a higher flexibility of the whole system but it will reduce the efficiency of the wind turbines due to the Maximum Power Point Tracking strategy cannot be used any more. The double battery energy storage system (BESS) strategy, which uses two parallel battery energy storage systems, could decouple the charging and discharging processes of batteries; however, the control of the conversion time is highly demanding. Besides, some other optimization methods have also proposed on the scheduling level.
In addition, in Chinese patent application No. CN102664422A, there is disclosed a method for smoothing the wind farm output which modifies the wind farm output by using wind power prediction and uses an improved adaptive filtering approach to filter the modified wind farm output.